Lab on Inference for the population mean



Lab on Inference for the population mean

(To review of Selected material from chapters 13,14,15 and 16)

Necessary data are available from the CD that comes with your book, but I have also uploaded them to Blackboard and the web page.

===================================================================

I) Breast feeding and bone loss

“Bone loss by nursing mothers. Breast-feeding mothers secrete calcium into their milk. Some of the calcium may come from their bones, so mothers may lose bone mineral. Researchers measure the percent change in mineral content of the spines of 47 mothers during three months of breast-feeding “.

1.1 What is the variable of interest in this study? ______________________

Is it ‘quantitative’ or ‘categorical’? _________________________________

Is this an observational of experimental study? _______________________

1.2 Assume that we know that the standard deviation of the variable ‘% change of bone mineral density in the spine’ of all nursing mothers is 2.5%. The mean of the 47 observations is -3.587

Calculate a 99% confidence interval for ‘mean % change of bone mineral density in the spine.

1.3 Write the value of the ‘margin of error’ in the previous question ___________

1.4 Imagine that you would like the margin of error to be only 1, keeping the confidence at 99%. How large should the sample be in order to achieve that? _____________

1.5 The formula you used to answer question 1.1 involves z* (value from the normal table), and is based on the fact that [pic] has a normal distribution. Why did you use that formula because…….

a) we know the value of [pic]

b) even when they did not tell us that the variable % change in mineral content has a normal distribution, the sample size is large enough as to claim the central limit theorem.

c) Both of the above

My answer is: ________

1.6 Now open the file boneloss.mtw from the webpage. Now assume that you do not know [pic] and you rely only on the observations, as is usually the case in real life.

a) Calculate the mean ____________and standard deviation __________in the sample

b) Calculate a 99% confidence interval for the population mean _________________

(You can use the formula or if you prefer you can calculate it using the computer: STAT>Basic Stat> 1 sample t, copy and paste the output here)

c) Interpret the confidence interval you obtained (don’t just repeat a definition, be specific in your interpretation according to the problem)

=================================================================

2. Observational studies suggest that moderate use of alcohol reduces heart attacks, and that red wine may special benefits. In an experiment, healthy mean were assigned at random to several groups. One group of 9 men drank a bottle of red wine each day for two weeks. The level of polyphenols (substances that affect cholesterol) in their blood was measured before and after the two-week period. The data (percent changes in level) are in (redwine.mtw)

a) Would you use the formula with z* or the formula with t* in this case? Why?

b) Why would we be worried about the variable ‘% change in polyphenol’ having a normal distribution?

c) Calculate the 90% confidence interval.

d) Interpret the confidence interval you calculated in c). Please be specific, it is not enough just to mention the numbers.

e) Somebody considers that confidence interval to be too wide. How large does the sample need to be in order to have a margin of error of only 1 %? (Since we do not know the value of [pic], we will use the current study as a preliminary study and assume [pic]=3 to be on the safe side) __________________

====================================================================

3) Students attitudes. The Survey of Study Habits and Attitudes (SSHA) is a psychological test that measures students’ study habits and attitude toward school. Scores range from 0 to 200. The mean score for all college students is 115 and the standard deviation is about 30. A teacher suspects that the mean for older students is higher than 115.

3.1) Write the null and alternative hypothesis (you might need the symbol [pic]so I put it here for you to copy and paste it when you need it).

Ho: _____________

Ha:_____________

3.2) Is this a ‘one sided alternative’ or a ‘two sided alternative’ ______________

Based on your answer about the type of alternative, the p-value would be an area in:

a) both tails of the normal distribution

b) just in the right tail

c) just in the left tail

My answer is _________

3.3) The teacher collects a simple random sample of 25 older students, gives them the test and calculates the sample mean of the scores which happens to be 125.8. You want to make your decision using [pic]=0.05

Calculate the test statistic ________________

Calculate the p-value ______________________

What is your decision about Ho? ________________

Which of the following options is a better interpretation of your decision? ___

a) Older students on average have a better attitude and study habits than the general population of older students

b) There is not enough evidence to conclude that older students on average have a better attitude and study habits than the general population of older students

3.4) Now we will repeat the problem but assuming that we don’t know that [pic]and so we need to work with the standard deviation of the sample instead, which happens to be s=25. The scores in the SSHA have an approximately normal distribution so we are not too worried about the necessary assumption. (You can use the computer to do the requested calculations and find the p-value.)

Calculate the test statistic ________________

Calculate the p-value ______________________

What is your decision about Ho? ________________

Which of the following options is a better interpretation of your decision? ___

c) Older students on average have a better attitude and study habits than the general population of older students

d) There is not enough evidence to conclude that older students on average have a better attitude and study habits than the general population of older students

3.5) Between which values you would say, with a 95% of confidence, that the mean for the SSHA scores for older students is? (Based on the fact that the mean for a random sample of 25 students was 125.8 and the standard deviation in the sample was 25)

______________________________

3.6) Assume you solved question 3.5 before question 3.4. Why you could have not just directly used your answer in question 3.5 to answer question 3.4?

___________________________________________________________

4) Generic vs. reference drugs. (genvsref.mtw)

• Control of other variables: all were male , healthy and nonsmoking and since each individual might respond differently , each individual was given both drugs

• Randomization: the order of application of the drugs was determined at random

• Replication: there are 20 individuals

This is a typical ‘match pairs’ case because each subject tried both drugs the generic and the reference, and the analysis is focused on the difference in the absorption of both drugs. The variable of interest is d=reference – generic.

The parameter of interest is in this case is [pic] (the mean difference between absorption of the reference and absorption of the generic).

4.1 The research question is ‘Do the two drugs differ significantly in absorption?

Write the null and the alternative hypotheses:

Ho: __________

Ha:___________

Is the alternative hypothesis ‘one sided’ or ‘two-sided’ ______________?

4.2 In order to use the t-procedures , especially since the sample size is moderately small, we need to check if the variable ‘difference’ does not show serious departures from normality.

Here are the histogram and the normal probability plot that indicate that departure from normality is not serious.

|[pic] |[pic] |

(genvsref.mtw)

Perform a matched pair t-test to test the hypotheses you wrote in question 4.1. You can either use the option STAT>BASIC STAT> Matched pair t test in Minitab (see screen below)

|[pic] |Make sure to indicate the type of alternative hypothesis in the |

| |options menu |

| | |

| | |

| |Minitab calculations are based in the formula |

| |[pic] |

| | |

| |And the t-student distribution with n-1 degrees of freedom to |

| |find the p-value. |

a) What is the value of the test statistic ‘t’ ? __________

What is the p-value? _______________

b) Assume we want to work with [pic]What is your conclusion ? Reject Ho or not reject Ho? ____________________

Why? __________________________________________________

c) Interpret your decision in b) by answering in plain English your answer to the research question ‘Do the reference and the generic drugs differ significantly in absorption?’

3. Calculate a 95% confidence interval for the mean difference in absorption between the reference and generic drug. _______________________

In this case, if you had had the answer to question 4.3 prior to working question 4.2, could you have answered parts b) and c) of question 4.2 without doing the calculations specified in part a) of question 4.2)? _______

Why is that in this case you could have used your answer to 4.3 to answer 4.2)? ________________

................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download